突出危险工作面瓦斯涌出异常识别与预警系统研究

发布时间：2018-04-22 11:45

本文选题：煤与瓦斯突出 + 时间序列　；参考：《中国矿业大学》2015年硕士论文

【摘要】：论文以原始瓦斯涌出监控数据为研究对象,以数据挖掘的思想对瓦斯涌出的异常状态曲线进行特征分析,结合小波阈值去噪的方法进行趋势分析,建立了基于小波阈值去噪的分级识别预警模型,并通过软件编程技术建立了煤与瓦斯突出识别预警系统。论文的主要研究内容包括以下几个方面:(1)瓦斯涌出时间序列的建立经过分析发现原始瓦斯数据时间间隔不等,存在缺失数据与异常数据;以1分钟为时间间隔、取分钟内瓦斯浓度平均值的方法建立瓦斯涌出时间序列,并通过取前后平均值的的方法进行补充与清理,建立了符合可比性原则的瓦斯涌出时间序列。(2)煤与瓦斯突出的识别预警通过对瓦斯涌出时间序列数字特征的分析,得出瓦斯涌出时间序列的一般性质;通过对瓦斯涌出时间序列的不同状态进行对比分析,得出存在突出危险性状态的特征;在此基础上结合小波阈值去噪的方法进行瓦斯涌出时间序列的动态趋势分析,最终建立基于小波阈值去噪的分级识别预警模型。(3)系统实现与验证采用Client/Server架构、Visual Studio开发平台、Microsoft SQL Server数据库,建立了基于具有可比性的瓦斯涌出时间序列、小波阈值去噪分级识别预警模型的煤与瓦斯突出识别预警系统,并结合实际煤矿瓦斯监控数据进行了验证。
[Abstract]:In this paper, the original monitoring data of gas emission is taken as the research object, the abnormal state curve of gas emission is analyzed by the idea of data mining, and the trend is analyzed by wavelet threshold de-noising method. The classification recognition and warning model based on wavelet threshold denoising is established, and the coal and gas outburst recognition and warning system is established by software programming technology. The main research contents of this paper include the following aspects: 1) the establishment of the gas emission time series. It is found that the original gas data have different time intervals, there are missing data and abnormal data, and the time interval is 1 minute. The time series of gas emission is established by taking the average value of gas concentration in minutes, and the gas emission time series is supplemented and cleaned by the method of taking the average value of gas concentration before and after taking, The identification and early warning of coal and gas outburst is established according to the principle of comparability. The general character of the time series of gas emission is obtained by analyzing the digital characteristics of the time series of gas emission. By comparing and analyzing the different states of gas emission time series, the characteristics of outburst dangerous state are obtained, and on this basis, the dynamic trend analysis of gas emission time series is carried out by combining wavelet threshold de-noising method. Finally, a hierarchical recognition and early warning model based on wavelet threshold denoising is established. The system is implemented and verified. The Client/Server framework is used to develop the Microsoft SQL Server database, and the time series of gas emission based on comparability are established. The early warning system of coal and gas outburst recognition based on wavelet threshold denoising and classifying recognition model is verified with actual coal mine gas monitoring data.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD713

【参考文献】